GTU Scientific Study Receives Best Paper Award at IEEE UBMK 2026

October 2, 2026 - GTU Office of Press and Public Relations
 

— An AI-supported research paper jointly prepared by GTU and Özyeğin University received the “Best Paper Award” at the 11th International Conference on Computer Science and Engineering (IEEE UBMK 2026), selected from more than 300 submissions.

— The joint study, which brings together materials science and deep learning and is supported by TÜBİTAK 1001, was selected as one of the seven best papers recognized by the Scientific Committee at IEEE UBMK 2026.

Gebze Technical University (GTU) achieved significant success at the 11th International Conference on Computer Science and Engineering (IEEE UBMK 2026), which brings together the latest developments, academic research, and industry applications in computer science and engineering on an international platform.

The paper titled “Obtaining High-Energy XRD Spectra from Low-Energy Measurements via Curriculum Learning on Gold Nanocrystals,” in which faculty members and students from our university participated as researchers, was selected as one of seven papers to receive the “Best Paper Award” following a rigorous evaluation by the Scientific Committee among more than 300 scientific papers presented at the conference.

The award-winning paper was jointly prepared by Prof. Yusuf Sinan Akgül, Dean of the Faculty of Engineering at Gebze Technical University; Berkay Serin, a graduate student in Computer Engineering at GTU; Erdem Nezihi Aksu, a fourth-year undergraduate student in Computer Engineering at GTU; and Assoc. Prof. Hande Öztürk from Özyeğin University.

A Computational Solution in Materials Science With Artificial Intelligence

X-ray diffraction (XRD), one of the fundamental methods used to analyze the atomic structure of materials, requires high-energy (short-wavelength) measurements for the highly reliable characterization of nanoscale crystals. However, such advanced measurements can only be performed at a limited number of large-scale research facilities worldwide, restricting laboratory access to these capabilities.

The award-winning study offers an innovative computational solution to this bottleneck through an artificial intelligence model. In the research, low-energy XRD measurement data, which are widely available in standard laboratories, are processed using curriculum learning and deep learning techniques, enabling the high-energy patterns that would otherwise be obtained through high-energy measurements to be predicted with high accuracy. The developed method aims to significantly reduce dependence on costly and difficult-to-access research infrastructure.

TÜBİTAK 1001 Support and Interdisciplinary Collaboration

This achievement was realized as an outcome of a comprehensive research project conducted in collaboration between Gebze Technical University and Özyeğin University and supported under the TÜBİTAK 1001 Scientific and Technological Research Projects Support Program. Bringing together materials science and artificial intelligence within a common methodology, the study demonstrates through a tangible achievement the integrative role of artificial intelligence in connecting different fields of expertise in multidisciplinary research.